1. Field of the Invention
This invention relates to a flow cell for measuring the color properties of a liquid, such as paint, having a transparent partition spaced from the window of the cell.
2. Description of the Art
Pigment dispersions and tints are widely used in formulating high performance liquid coating compositions. Such compositions are used, for example, as exterior finish paints for automobiles and trucks.
Dry color measurement of such liquid compositions is believed to be the most accurate indication of the composition's color properties. Such measurement is usually made manually by taking an aliquot of the composition being prepared. The composition is sprayed as a coating onto a panel and the panel is baked and dried. One or more color properties of the dried coating may be measured against a reference using a colorimeter or spectrophotometer. Based upon the measurement the batch under preparation is adjusted in an effort to obtain a closer match to the reference. Manual color measurements are very time consuming, primarily due to the long preparation and drying times. Also, the procedure may have to be repeated numerous times before the desired color property is achieved.
It is believed that manufacturing efficiencies may be achieved through the ability to measure the color properties of a liquid composition while in a wet state. However, to be effective, any wet color measurement must accurately predict the color of the composition when dried. This goal has proved elusive.
Instruments employing a reflectance spectrophotometer have been used to obtain a free surface reflectance measurement of a wet liquid dispersion. Representative of these instruments are the devices described in U.S. Pat. No. 6,583,878 (Hustert), U.S. Pat. No. 6,292,264 (Voye et al.) and German Patent DE 25 25 701 (Langer). These instruments all employ a free surface reflectance measurement of a wet coating film utilizing a spectrophotometer. The measurements taken from these instruments thus embody the best representation of the color of the coating film that could be correlated with the measurements of the same film in its dry state. However, surface non-uniformities of such wet coatings, as well as viscosity variations, settling, and flocculation could still lead to erroneous results and unacceptable measurement variability.
It is believed that further efficiencies can be achieved by coupling such a device to a manufacturing process. However, coupling such devices as just described to a continuous process has its own encumbering difficulties, including but not limited to, operation of said device in the presence of volatile flammable solvents emitted from the sample surface as well as cleaning.
To couple a color measurement device to a manufacturing process, in light of the aforementioned possible presence of volatile flammable solvents, as well as taking into consideration that many processes operate at super-ambient pressures, it is standard practice to contain the fluid sample flow through the device in a closed system, separated from the illumination source and spectral detector by a window of sufficient strength, and therefore thickness, to withstand said pressure. The thickness T required of such a window is given by the equation:
  T  =                    z        ⁢                                  ⁢        P        ⁢                                  ⁢                  D                      ′            ⁢                                                  ⁢            2                              σ      
where                z is a shape factor for the window;        P is the pressure being contained;        D′ is the unsupported diameter, and        σ is the maximum design stress (pressure) for the window material.        
Instruments which measure the absorbance and/or scattering properties of a liquid contained in a closed system have been proposed for standard spectrophotometric measurements, including both laboratory and process applications, either in transmission or reflectance mode. Some of these instruments also purport to measure the color of the liquid in reflectance mode through a sight glass into the process stream or within a sample cell employing a window between the sample and the detector. U.S. Pat. No. 4,511,251 (Falcoff et al.) and U.S. Pat. No. 6,288,783 (Auad et al.) are representative of this class of instrument.
The instrument described in the last referenced patent employs a variable pathlength measurement cell to measure properties of liquids, including color. The instrument employs a closed path for the flow of the liquid to be measured, thus allowing it to be placed in hazardous classification areas within a manufacturing plant environment. However, this particular instrument has multiple moving parts which are part of the liquid path, which can cause difficulty in cleaning, and are difficult to maintain. Another disadvantage is that the instrument requires high volumes of liquid sample to take proper readings. Moreover, while the instrument can measure in both reflectance and transmission modes, it employs 0/0 geometry for each. As a result, in transmission mode no information is provided about scattered light from the fluid being analyzed. In reflection mode unmitigated backscattered light from the source washes out the color sensitivity.
Ultimately, the single most significant issue to overcome in the measurement of the color of a liquid in intimate contact with the window of the flow cell is the disruption of light on its way back to the detector that occurs because of the presence of the window itself. Causes of such disruption of the light include, but are not limited to, reflection, refraction, total internal reflection, and loss or escape of said light with reference to the various surfaces of the window. As a result of such disruption the light ultimately either never reaches the detector or is modified by the surfaces of the window with which it interacts, such that spectral information presented to the detector is no longer truly representative of the sample being measured.
A liquid in intimate contact with a viewing window looks different to the human eye when viewed through that window than the color of the same liquid when viewed in a free surface fashion, i.e., with nothing between the eye and the free surface of the wet liquid.
FIG. 1 is a stylized diagrammatic representation of the optical phenomena occurring at the interface between a liquid L and a window W. The window W may form part of a flow cell or a probe. The liquid L is flowing past the window in a flow direction G at some predetermined fluid pressure. The liquid L is in contact with the window W. The light scattering pigments of the liquid composition are usually dispersed in a solvent vehicle that has an index of refraction close to the index of refraction of the window material.
To gain a better understanding of the optical effects that occur when a liquid is viewed through a window, consider the situation depicted in FIG. 1. As a light ray R propagates through a medium M (e.g., air) it impinges upon the exterior surface E of the window W. The material of the window W refracts the ray R. The refracted ray R′ propagates through the window W toward the window/liquid interface. If the indices of refraction of the window and the solvent are substantially equal (i.e. within about 0.2 refractive index units of each other) no optical interface exists between the liquid and the window and the ray continues along substantially the same path.
The light ray R′ that enters the liquid and strikes a suspended pigment particle is both specularly reflected and diffusely scattered into a solid hemisphere of 2π radians emanating from a scatter site X. (It is noted that although the scattering occurs within the liquid the scatter site X is illustrated in FIG. 1 at the window/liquid interface). The scattered specular rays, e.g., the ray S, impinges against the window surface E at an angle θS (measured with respect to a normal to that surface) that is less than the critical angle θc of the window/medium interface. Such a scattered specular ray S exits the window (at point Q) into the field of view F presented to a detector.
However, some diffusely scattered rays, e.g. the ray U, which emanate from the scatter site X, impinge against the window surface E at an angle θU that is greater than the critical angle θc. Such a diffusely scattered ray U is totally internally reflected within the window (at point V). The diffusely scattered ray U propagates back toward the window/liquid interface where it may undergo a secondary scattering impact at site X′, at which point its scattering angle may change direction.
The secondary scattering impact at site X′ itself produces specular and diffuse scatterings. Such a scenario is repeated several times within the window material. At each scattering impact some of the light is reflected at angles which would render its direction at the window surface E greater than the critical angle for the window/air interface while some of the light is reflected at angles which would render its direction at the window surface E less than the critical angle for the window/air interface.
The distance d between the initial impact site X and a secondary impact site X′ depends on the thickness T of the window W according to the relationship:d=2·T tan θu,                where θu is the angle that the diffusely scattered ray U makes with the normal to the surface E.        
Owing to the fact that, as discussed earlier, the window must be thick enough to withstand the pressure of the sample stream it may be the case that there is insufficient lateral distance available for a diffusely scattered ray U to undergo a statistically significant number of secondary impacts before being scattered at an angle with respect to the normal to the surface E that is less than the critical angle for the window/air interface. In that case the ray U is more likely to exit through the peripheral surface P of the window W, as indicated at point Z. This energy is outside of the field of view F and is lost to the detector.
The effect caused by total internal reflection of diffusely scattered rays is twofold. Firstly, the intensity of the scattered light ultimately reaching the detector is diminished. This makes the liquid appear darker in color. Secondly, total internal reflection causes the body of the window to exhibit a “glow” effect. This increases the background against which detected radiation is measured.
The diminution in received intensity coupled with an increase in background intensity produces a flattening of the waveform of the intensity/wavelength curve or detected reflectance spectrum. When standard colorimetric calculations are carried out to calculate L*, a* and b* according to the CIELab76 formalism, the net effect of this is to produce a loss of chroma (C*ab=[a*2+b*2]1/2), and to skew the determination of perceived color properties. Moreover, since the intensity undergoes different range distortions in different localized wavelength domains, the problem cannot be expeditiously cured by merely scaling the resulting intensity waveform. Furthermore, if the light is disrupted on its way back to the detector in a way that misrepresents measurement of the true color of the sample, it follows that making adjustments to that color, such as may be required in a manufacturing process, may also be in error.
Accordingly, in view of the foregoing it is believed advantageous to provide an apparatus and a method which mitigates the disruption of light, and hence the loss of chroma, during color measurement of a liquid material using reflectance spectroscopy. It is also believed advantageous that such liquid measurements correlate well to measurements made on the material in its dry state.
It is believed to be of further advantage that the apparatus and method be able to operate in the environment of a pressurized liquid without alteration of the color measurement.
It is believed to be of still further advantage to provide an apparatus where pressurized liquid is introduced into a measurement region without undergoing any flow discontinuity so that a laminar flow of pressurized liquid flow is maintained past the window.
It is believed to be of yet further advantage to provide an apparatus that is able to be cleaned rapidly (e.g., within one or two minutes) so that the cycle time of the measurement is extremely small compared to process changes; that affords easy (including automatic) delivery of a sample to the analysis cell so that measurements of color can be made rapidly; and which can be placed in a potentially hazardous environment, such as a plant floor.